[docs] Details to performance_roc()

[DominiqueMakowski]

There is probably a better way to phrase it though (@mattansb?)

[mattansb]

The docs look great (:

[DominiqueMakowski]

why does the as.numeric() method fail?? It fails locally too, did I forgot how to R?

[strengejacke]

For me, it also fails locally and I couldn't find any mistake? Strange, what have we missed?

[strengejacke]

performance:::as.numeric.performance_roc(x) works, as.numeric(x) not. @etiennebacher, any ideas, why the as.numeric.performance_roc() method in this PR does not work, resp. is not called?

library(performance)
library(bayestestR)
data(iris)

set.seed(123)
iris$y <- rbinom(nrow(iris), size = 1, .3)
folds <- sample(nrow(iris), size = nrow(iris) / 8, replace = FALSE)
test_data <- iris[folds, ]
train_data <- iris[-folds, ]

model <- glm(y ~ Sepal.Length + Sepal.Width, data = train_data, family = "binomial")
x <- performance_roc(model, new_data = test_data)

class(x)
#> [1] "performance_roc"     "see_performance_roc" "data.frame"

methods(class = class(x))
#> Warning in .S3methods(generic.function, class, envir, all.names = all.names, :
#> 'class' is of length > 1; only the first element will be used
#> [1] as.numeric plot       print     
#> see '?methods' for accessing help and source code

as.numeric(x)
#> Error: 'list' object cannot be coerced to type 'double'

performance:::as.numeric.performance_roc(x)
#> [1] 0.3766234

^{Created on 2024-12-29 with reprex v2.1.1}

[etiennebacher]

You need to add the S3 method for as.double.performance_roc instead of as.numeric.performance_roc. From ?as.numeric:

as.numeric is a generic function, but S3 methods must be written for as.double. It is identical to as.double.

[strengejacke]

Coo, thanks!